Developing photoresist PCB with UV bank note checker

Question

I'm just looking into developing a photoresist PCB. I bought a kit that contains a photoresist single sided PCB, developer solution (in powder form to mix with water), etchant, and a tank.

I have printed my circuit diagram on an inkjet transparency.

The only thing I was missing was a UV light to develop the photoresist. I tried it with a standard lamp and had no results after ~20 minutes, so I figured I needed a genuine UV light. It's too dark in the evenings this time of year for me to be able to do it outside.

I am looking into buying one of those bank note checkers that shops use to check for counterfeit money. It's a downward facing 4W bulb - here:

http://www.maplin.co.uk/p/mains-powered-uv-fluorescent-light-source-mw36p

Will this be sufficient to develop my photoresist? I don't need it done in a hurry and don't mind waiting a long time.

Update: Well I'm happy to confirm I got decent results from using the note checker, OHP inkjet transparencies, and some positive photoresist from Maplin. I just did a test, around 8 minutes under the lamp, then I got called for dinner! So about half an hour later, dunked it in a reasonably weak developer solution - maybe 100ml of water and two or three sprinkles of Sodium Hydroxide.

As you can see, the one closest to the camera turned out well (the scrappy bit on the side is because I cut the board with tin snips - not a great idea - and had a bit of the protective film left which I peeled away. It made for a good comparison). It's a bit weak in places, but those square pads on the right are perfect. The one behind it, not so much. I think I overexposed that one. It was the cheap board that came with the kit.

Answer 1

The rule is not UV = UV = UV. UV only means the waves are shorter than what someone at some point called violet light.

If you look at this Material Sheet for a spray on Photo Resist, while it's not a great sheet, it's the first I found typing in photoresist on Farnell.

It says "The Spectral Sensitivity of the photoresist is between 350 and 400 nanometers."

A lamp you want to use should fall in that range, for this resist. Many resists have similar windows, I happen to know one I used to use had a peak specifically at 358nm, but I have seen ones that go up to 420nm.

A simple money Checker usually uses normal TL coated with filter, which puts it at 405-ish nm. If it's a cheap one, who knows for sure, really. 405nm would be too high for me to be willing to buy and try, certainly with only one attempt available to me.

---

This brings me to another warning I want to give you: Home printer transparencies and home printer ink notoriously is not awesome at UV. If you have no stats there's a chance of the plastic being bad at allowing light below 400nm through, though that'd not be my primary worry. There's also every chance of the inkjet ink not blocking the light well enough, this would be more of a worry. Back in the day someone at a large Company here in Eindhoven would help me make my transparencies on a professional thick-layer toner-based process, but he still advised me to put two on top of each other. (Though admittedly using a particularly high-intensity light source too)

I guess what I'm saying is, try a simple block/line/triangle mix of open and closed transparency on a small corner before you fudge up your entire experiment. Especially if you don't know the exact optical power at certain wavelengths of your light source and/or can't calculate the appropriate exposure time to withing 10%~20%, experimenting with a set-up is always the better way to go before you dive in. In a professional system many of these parameters are known much more accurately and even then with new systems they often need a few attempts to get all the settings spot-on.

Answer 2

Using a UV note checker from Maplin, Inkjet Transparencies and photopositive board, I put the transparency on the board and put it under the lamp at close range (tilting the lamp forward slightly) and left it there for around 8 minutes.

When put in the developer solution around half an hour later, I had a good result. Some lines were a little weak but this could be down to the thirty minute wait.